Skip to main content

Full text of "A New Species Of Microgale (Lipotyphla : Tenrecidae : Oryzorictinae) From The Foret Des Mikea Of Southwestern Madagascar"

See other formats

1 17(3):25I-265. 2004. 

A new species of Microgale (Lipotyphla: Tenrecidae: Oryzorictinae) 
from the Foret des Mikea of southwestern Madagascar 

Steven M. Goodman and Voahangy Soarimalala 

(SMG) Department of Zoology, Field Museum of Natural History, 1400 Roosevelt Road, 
Chicago, Illinois 60605, U.S.A., e-mail: and WWF-Madagascar, BP 738, 
Antananarivo (101), Madagascar, e-mail:; 

(VS) Departement de Biologie Animale, Universite d’ Antananarivo, BP 906, Antananarivo (101), 
Madagascar and Ecology Training Program, WWF-Madagascar, BP 738, Antananarivo (101), 
Madagascar, e-mail: 

Abstract. — A new species of Microgale, M. jenkinsae (Lipotyphla: Tenre- 
cidae), is described based on two specimens taken during an early 2003 bio- 
logical survey of the Foret des Mikea in southwestern Madagascar. It is distin- 
guished from other congeners by numerous pelage, cranial, and dental char- 
acters. M. jenkinsae is the fourth known species in this genus confirmed to 
occur in the dry western and southern forests of the island. The Foret des 
Mikea, the only site M. jenkinsae is known from, is the last remaining block 
of a distinctive forest habitat and is under considerable threat from human 
habitat degradation. Action needs to be taken to protect this unique region. 

Resume. — Une nouvelle espece de Microgale, M. jenkinsae (Lipotyphla: 
Tenrecidae), est decrite a partir de deux specimens recoltes au cours d’un in- 
ventaire biologique mene au debut de l’annee 2003 dans la foret des Mikea 
situee au sud-ouest de Madagascar. On le distingue de ses autres congeneres 
par divers caracteres de pelage, craniens et dentaires. M. jenkinsae est la qua- 
trieme espece connue de ce genre dont la presence est confirmee dans les forets 
seches de l’ouest et du sud de 1’ile. La foret des Mikea, le seul site d’ou M. 
jenkinsae a ete rapporte, est le dernier bloc d’un habitat forestier distinctif qui 
est cependant extremement menace par la degradation de F habitat perpetree 
par l’homme. Des actions doivent etre prises pour proteger cette region unique. 

On the island of Madagascar there is an 
endemic family of Lipotyphla, known as 
the Tenrecidae, that represents one of the 
most remarkable adaptive radiations found 
in living mammals (Olson & Goodman 
2003). As currently circumscribed. Micro- 
gale (shrew tenrecs) a tenrecid genus in the 
subfamily Oryzorictinae, comprises 1 8 spe- 
cies (Jenkins 2003). On the basis of biolog- 
ical inventories and associated museum 
studies conducted over the past few de- 
cades, seven species of Microgale new to 
science have been named (Jenkins 1988, 
1992, 1993; Jenkins et al. 1996, 1997; 
Goodman & Jenkins 1998; Jenkins & 

Goodman 1 999), although one of these, M. 
pulla, has since been synonymized (Jenkins 
et al. 1997). Subsequent to the publication 
of MacPhee’s (1987) taxonomic revision of 
the genus Microgale, there has been a re- 
newed interest in the small mammals of 
Madagascar. With the advent of pit-fall de- 
vices to trap these animals, there has been 
a massive increase in available shrew tenrec 
specimens. This has lead to a series of pub- 
lications refining some of MacPhee’s taxo- 
nomic conclusions and a greater under- 
standing of intra-specific, particularly age 
related, and inter-specific variation amongst 
these animals. As witnessed by the recent 



Fig. 1. Map showing the collection locality of Microgale jenkinsae in the Foret des Mikea, the forested 
zone between Morombe and Manombo, in southwestern Madagascar (expanded inset). 

description of six new valid shrew tenrec 
taxa (an increase of 33%), Microgale tax- 
onomy is in flux as a result of ongoing bi- 
ological inventories and molecular and 
morphological studies. 

Of the currently recognized 1 8 species of 
Microgale, 15 are restricted to the eastern 
and northern moister portions of Madagas- 
car where they occur in either forests or 
marshes. Of the remaining three species, 
two have been collected over the past few 
decades in the dry western forests. These 
include M. brevicaudata, which occurs 
from the northern foothills of the Marojejy 
Massif in the northeast, a zone of humid 
forest but probably with a marked dry sea- 
son, north to Vohemar and the region of 
Antsiranana at the north end of Madagascar, 
and then south along the west portion of the 

island to at least the Onilahy River near To- 
liara (Goodman et al. unpublished; Fig. 1). 
The second species, M. nasoloi, is known 
from two inland isolated forests in south- 
western Madagascar in the vicinity of Sa- 
karaha — the Analavelona Massif and the 
Foret de Vohibasia (Jenkins & Goodman 
1999; Fig. 1). M. longicaudata is the third 
species falling into this group and has been 
collected from both eastern humid forests 
and western dry forests. However, M. lon- 
gicaudata, as currently defined, includes 
several cryptic species and will soon be re- 
vised (Olson et al. in press). 

MacPhee’s (1987, Fig. 13) map of col- 
lecting localities for Microgale included 3 1 
sites in the eastern humid forest where a 
total of nine species were trapped, two sites 
in the western dry deciduous each with sin- 



gle species (one based on owl pellets), and 
three sites in the southern spiny-bush each 
with two species (all based on owl pellets). 
Although considerable advances have been 
made concerning the species richness and 
distribution of shrew tenrecs since Mac- 
Phee’s important revision, these data indi- 
cate a greater diversity of this group in the 
more mesic portions of the island. Recent 
biological inventories of the western and 
southern forests of Madagascar have largely 
upheld this view. However, during a 2003 
survey of the Foret des Mikea, the region 
between Morombe and Manombo (Fig. 1), 
we captured a Microgale that represents a 
previously undescribed species of shrew 

Materials and Methods 

Our small mammal collection made in 
the Foret des Mikea contains two speci- 
mens of Microgale, and in order to deter- 
mine their taxonomic identity, we have con- 
sulted material housed in several natural 
history museums, which include: BMNH — 
The Natural History Museum, London (for- 
merly British Museum of Natural History); 
FMNH — Field Museum of Natural Histo- 
ry, Chicago; MNHN — Museum National 
d’Histoire Naturelle, Paris; and UADBA — 
Universite d’ Antananarivo, Departement de 
Biologie Animale. 

Five external measurements in millime- 
ters were taken from our two specimens be- 
fore preparation and included: total length, 
head and body length, tail length, hind foot 
length (not including claw), and ear length. 
Mass was measured with the use of a spring 
balance and recorded in grams. 

An additional six cranial and two dental 
measurements were taken using a digital 
calipers accurate to the nearest 0.1 mm. 
These measurements, and their definitions, 
are: breadth of braincase: the greatest dis- 
tance measured across the hamular process- 
es of the squamosals to the mastoid bullae; 
greatest length of skull: the distance be- 
tween the tips of the nasals and the poste- 

rior most portion of the cranium; interor- 
bital breadth: the minimum distance 
across the frontal bones between the orbital 
fossae; length of mandibular tooth row: 
the maximum distance from distal surface 
of the third molar to anterior surface of the 
first incisor; length of nasal: the maximum 
distance from the posterior extension of the 
nasals to their anterior tip; length of palate: 
the shortest distance between the tip of the 
postpalatal spine and anterior surface of the 
first upper incisor; length of maxillary 
tooth row: the maximum distance from dis- 
tal surface of the third molar to anterior sur- 
face of the first incisor; and zygomatic 
breadth: the maximum span between the 
zygomatic processes of the maxillae. 

Tooth abbreviations include: I 3 incisor, 
d = deciduous, C = canine, PM = pre- 
molar, and M = molar. Upper case tooth 
abbreviations with superscript are used for 
upper teeth and lower case abbreviations 
with subscript for lower teeth. Cranial and 
dental nomenclature follows Hershkovitz 
(1977) and MacPhee (1987). 

After comparison of the two specimens 
collected in the Foret des Mikea to all de- 
scribed forms of Microgale, these individ- 
uals could not be allocated to any known 
form and are therefore described as a new 

Microgale jenkinsae, new species 
Fig. 2, 3, Tables 1, 2 

Holotype . — FMNH 176215, sub-adult 
male, collected on 18 February 2003 by 
Steven M. Goodman and Voahangy Soari- 
malala, field number SMG 13489. The 
specimen was preserved as a round study 
skin, with associated skull and partial post- 
cranial skeleton. Tissue samples were pre- 
served in EDTA. The skin is in good con- 
dition with a small hole in the left thigh. 
The skull and partial postcranial skeleton 
are intact. Dental age is sub-adult with I 3 
still erupting and matches MacPhee’s 
(1987) eruption pattern stage 1. The basis- 
penoid-basioccipital sutures are unfused. 



Fig. 2. Photograph of the holotype of Microgale jenkinsae (FMNH 176215), a sub-adult male collected on 
18 February 2003 in the Foret des Mikea, 9.5 km west Ankiloaka, 22°46.7’S, 43°31.4’E. (Photograph taken by 
S. M. Goodman.) 

Table 1 . — External measurements (in millimeters) and weight (in grams) of Microgale jenkinsae and other species of small Microgale. Measurements presented as 
mean ± standard deviation (minimum-maximum, n). For samples of two or fewer specimens only the measurements are presented. 



cb oi vo vd 

vo vo 

^ ~ w g w ^ ^ 

External measurements are: total length 
1 43 mm, head and body length 62 mm, tail 
length 79 mm, hind foot length (without 
claw) 15 mm, and ear length 18 mm. The 
animal weighed 4.9 gm (Table 1). 

Type locality . — Madagascar: Province de 
Toliara, Foret des Mikea, 975 km west An- 
kiloaka, 22°46.7'S, 43°31.4'E, elevation 
about 80 m above sea level (Fig. 1). The 
site is about 17 km inland from the Moz- 
ambique Channel. 

Habitat . — The holotype was obtained in 
partially disturbed dry transitional decidu- 
ous forest growing on red sands. It was cap- 
tured in a pitfall trap placed in relatively 
dense understory composed primarily of 
Xerophyta (Velloziaceae), small bushes, 
and succulent Euphorbiaceae. 

Diagnosis. — A relatively small member 
of the genus Microgale with a head and 
body length of 59-62 mm, tail length of 
79-81 mm, and greatest skull length of 
18.7-18.8 mm. Deciduous PM 2 is simple, 
caniniform, and single-rooted. The color of 
the dorsal pelage is mixed agouti and the 
venter is gray with white-tipped fur. The 
ears are notably long (18 mm) for a shrew 
tenrec of this size. 

Paratype . — FMNH 176154 (SMG 
13492), sub-adult female from the same lo- 
cality as the holotype, collected 19 Febru- 
ary 2003, and prepared as fluid preserved 
specimen with extracted skull. The dental 
eruption pattern fits MacPhee’s (1987) stage 
1. Tissues saved in EDTA. 

Distribution. — Microgale jenkinsae is 
known only from the type locality in the 
Foret des Mikea, southwestern Madagascar. 
Description . — A small species of Micro- 
gale having a tail longer than the head and 
body (Fig. 2). The dorsal fur is relatively 
dense and soft. Pelage from the level of the 
ears to the base of the tail (including the 
flanks), is a mixture of completely black 
and tannish-brown hairs, or those that are 
tannish-brown along most of their length 
and black-tipped, imparting an agouti ap- 
pearance. The agouti pattern runs anteriorly 
from the level of the ears to the eyes. An- 



Table 2. — Cranial and dental measurements (in millimeters) and weight (in grams) of Microgale jenkinsae 
id other species of small Microgale. Measurements presented as mean ± standard deviation (minimum — 
laximum, n). For samples of two or less specimens only the measurements are presented. 


Gre » tH 

z Ef 


M. jenkinsae 




(Holotype FMNH 176215) 

M. jenkinsae 




(FMNH 176154) 

M. nasoloi 




(Holotype FMNH 156187) 

M. pusilla 

16.6 ± 0.71 

6.1 ± 0.24 

3.4 ± 0.17 

15.7-17.5, n = 7 

5. 6-6. 3, n = 8 

3. 1-3.7, n = 9 

M. parvula 

16.5 ± 0.47 

5.1 ± 0.23 

3.7 ± 0.19 

15.5-17.0, it 12 

4. 7-5. 4, n = 12 

3. 3-4.0, n = 12 

M. brevicaudala 

20.7 ± 0.79 

8.0 ± 0.51 

4.8 ± 0.27 

19.0-21.9, n = 13 

7.2— 8.8, n)^(13 

4.3— 5.2, n = 14 


21.1 * 0.76 

8.2 ± 0.32 

5.0 ± 0.18 

20-22, n = 12 

7. 7-8. 7, n = 1 1 

4.6-5. 2, n = 12 

terior and lateral to this band, the pelage is 
distinctly paler in coloration, with the ma- 
jority of hairs being pale tan to silvery- 
white. Individual hairs along the dorsum 
measure 4-5 mm. Guard hairs are medium 
gray in color. The ventral pelage, with the 
exception of the portion surrounding the 
gular to mental regions, is gray based with 
off-white tips. The difference between the 
ventral and dorsal color pattern is pro- 
nounced, but grade into each other laterally 
instead of forming a well-demarcated line. 
Upper surfaces of fore feet and hind feet 
are covered with short silver-white fur, 
which on the hind feet extends slightly be- 
yond the claws as ungual tufts. The color 
of mystacial and rhinarial vibrissae vary 
from either completely beige-white or 
black, to black at the base and gradually 
becoming beige-white at the tips. Mystacial 
vibrissae reaching up to 20 mm and rhinar- 
ial vibrissae about 5 mm in length. Pinnae 
are notably long (18 mm) for a small Mi- 
crogale, dark brown in color, and covered 
internally and externally with fine, silvery- 
gray fur. 

The hind foot is relatively long (14-15 
mm) for a small species of Microgale (Ta- 
ble 1). The first digit of the hind foot is less 
than one-third the length of the second dig- 

it. The second and third digits are subequal 
in length, with the fourth digit slightly lon- 
ger. The fifth digit is about two-thirds the 
length of the fourth. There are five plantar 
tubercles and, based on FMNH 176154 
(SMG 13492), these are located at the base 
of digit 1 and digit 5, in intermediate po- 
sitions between the base of digits 2 and 3 
and digits 3 and 4, and notably reduced as 
distal hypothenar and proximal thenar pads. 

The skin of the tail is dark brown dor- 
sally and tannish-brown ventrally, and 
forming a relatively well-demarcated line 
laterally separating these two surfaces. The 
tail is clothed with very fine silvery-white 
fur, which becomes slightly denser at the 
tip. In FMNH 176154 (SMG 13492) the 
last 10 mm of the tail is mottled dark- white. 

The skull is relatively short (Table 2), 
slightly flattened dorsolaterally, with a con- 
stricted interorbital region. The rostrum is 
relatively short and tapers anteriorly. The 
anterior portion of frontals consist of two 
slightly concave plates divided at the mid- 
dorsal line and the posterior portion is 
slightly-domed. The braincase has a slightly 
bulbous parietal and interparietal, a rounded 
supraoccipital and occipital, and a weakly 
defined occipital crest. Dentally the holo- 
type is a sub-adult with the erupting crown 



Table 2. — Extended. 



Length of 


of nasal 

maxillary toothrow 

mandibular toothrow 
















6.9 ± 0.11 

6.8- 7. 1, n = 7 
6.8 ± 0.18 

6.5— 7.0, n - 12 
8.7 ± 0.27 
8.2-9. 1, n = 13 
9.3 ft 0.26 

8. 9- 9.7, n = 12 

7.0 ± 0.37 

6.4— 7.5, n = 8 

7.3 ± 0.22 

6. 9- 7. 7, n = 12 

9.3 ± 0.62 

8.4- 10.5, n — 14 

9.4 ± 0.45 

8.9- 10.3, n gS|2 

7.3 ± 0.34 

6.8— 7.8, n = 8 

7.3 ± 0.23 

6.9— 7.6 , n = 1 1 

9.3 ± 0.57 
8.2-10.2, n = 14 

9.8 ± 0.39 
9.1-10.2, n= 12 

7.4 ± 0.40 

6.7— 7.7, n = 8 
7.4 ± 0.28 

6. 7- 7. 6, n - 12 
9.3 ± 0.51 

8.3- 10.2, n sf.14 
10.0 ± 0.43 

9.3- 10.7, n =1)2 

7.0 ± 0.36 
6.5-7 .4, n g 7 
7.2 ± 0.19 

7. 0- 7. 6, n = 12 
8.8 ± 0.35 

8. 1- 9.4, n = 14 
9.6 ± 0.47 

9.0-10.5, ne 12 

of I 3 present and alLantemolars are decid- 
uous, fitting stage 1 in MacPhee’s (1987) 
tooth eruption pattern. The upper toothrows 
from dl 1 to dPM 2 slightly converge anteri- 
orly. The lingual margins of dPM 3 and M 1 
to M 3 are roughly parallel. Palatal foramina 
are present. Pterygoids are relatively short 
and broad, and the pterygoid processes 
winged-shape and curved mid-ventrally. 
The glenoid fossa is shallow and narrowly 
curved. The mandibles are slender, the cor- 
onoid processes are relatively narrow at 
their bases and pointed dorsally, and the an- 
gular processes are short and narrow, and 
the dorsal surface is not expanded (Fig. 3). 

The dentition is not markedly robust 
(Fig. 3). There is a gap between the dl 1 and 
dl 2 and between the dpm 2 and dpm 3 . The 
first upper incisor (dl 1 ) is small, bicuspid 
(bidentate), and the distostyle moderately 
well developed; the dl 2 with approximately 
the same crown height as dl 1 , the tricuspid 
(tridentate) has the anterior accessory cusp 
more developed than distostyle; the dl 3 is 
one-half crown the height of dl 2 and reach- 
ing just beyond the level of the distostyle 
of the dl 2 , the bicuspid with small distos- 
tyle; the dC 1 robust with crown height 
reaching that of the dl 2 , with small acces- 
sory anterior cusp and pronounced distos- 

tyle; the dP 2 small, equal in crown height 
to distostyle of the dC 1 ; the dP 3 is large, 
slightly greater in crown height than the 
dC 1 , lingual ledge with well-developed pro- 
tocone, and the parastyle, mesiostyle, an- 
terior ectostyle, and distostyle present; the 
dP 4 is large, longer in crown length than M 1 
to M 3 , with elongated paracone, the lingual 
ledge with a protocone more developed 
than M 1 to M 3 , anterior ectostyle approxi- 
mately same length as paracone, and the 
parastyle, mesiostyle, and distostyle pre- 
sent; the M 1 and M 2 large, parastyle, me- 
siostyle, anterior ectostyle, and distostyle 
present, and centro-buccal cleft slightly 
more prominent in M 2 than M 1 ; the M 3 is 
reduced in size and compressed anteriopos- 
teriorly. The first lower incisor (di,) is large, 
slightly shorter in crown length to the di 2 , 
the posterior accessory cusp well-devel- 
oped; the di 2 is large, the posterior acces- 
sory cuspid well-developed; the di 3 is small, 
about one-half the crown length of lower 
(deciduous) canine; the dc is large, poste- 
rior accessory cuspid present, no anterior 
accessory cuspid; the dpm 2 is small, slightly 
shorter that posterior accessory cuspid of 
lower canine, poorly developed anterior ac- 
cessory cuspid and posterior accessory cus- 
pid, and single-rooted; the dpm 3 is moder- 



Fig. 3. Views of the cranium and mandible of the holotype of Microgale jenkinsae (FMNH 176215): upper 
left, dorsal view; upper right, ventral view; lower center lateral view of cranium and mandible. (Photograph 
taken by J. Weinstein, image number Z94379_05d.) 



ate in size, slightly longer in crown height 
than the pm 2 , moderately developed ante- 
rior accessory cuspid and posterior acces- 
sory cuspid; the dpm 4 is large, equal in 
crown height (formed by prominent proto- 
conid) to nq, the anterior accessory cuspid 
and posterior accessory cuspid present; the 
nrqand m 2 are large, the m, slightly subequal 
in crown height to the m 2 , both with well- 
developed protoconid, anterior accessory 
cuspid, and posterior accessory cuspid, and 
slightly elongated anterobuccal cingulum; 
the m 3 is large and equal to m 3 in crown 
height, and with a well-developed proto- 
conid, anterior accessory cuspid, and prom- 
inent posterior accessory cuspid, and slight- 
ly elongated anterobuccal cingulum. Given 
that the individuals of M. jenkinsae are 
stage 1 sub-adults, no information can be 
provided on the adult dentition or antemolar 
replacement pattern of this species. 

Comparisons. — The fact that our two 
specimens of Microgale jenkinsae are stage 
1 sub-adults complicates comparisons to a 
certain degree. However, sub-adult mem- 
bers of this genus, at this stage of dental 
eruption, exhibit the pelage coloration of 
adults and, in general are similar to adults 
in external measurements (MacPhee 1987; 
Jenkins et al. 1996, 1997). 

M. jenkinsae is readily distinguished ex- 
ternally from other relatively small mem- 
bers of this genus by pelage coloration and 
measurements. The contrasting agouti dor- 
sum and grizzled-gray venter is unique 
among small shrew-tenrecs. The pelage 
pigmentation in M. nasoloi is a relatively 
uniform gray; M. fotsifotsy has less gray in 
the dorsum than M. jenkinsae and is notably 
darker; M. parvula has a dark brown dor- 
sum and dark grayish-brown ventrum; M. 
brevicaudata is medium-brown dorsally 
and dull grayish-brown ventrally; and in M. 
longicaudata and M. pusilla the dorsum is 
a mixed light brown and medium brown 
and ventrum gray broadly edged with dark 
tan-brown. Further, there is no overlap in 
tail measurements between M. jenkinsae 
and any of these taxa, with the exception of 

M. fotsifotsy, but M. jenkinsae can be dif- 
ferentiated from it based on pelage charac- 
teristics, a non-white-tipped tail, upper sur- 
faces of the feet clothed with short silver- 
white fur, and several external measure- 
ments (Table 1). 

The presence of a single-rooted second 
lower premolar separates M. jenkinsae from 
all other named small members of the genus 
Microgale, with the exception of M. pusilla. 
In the latter species the root form is iden- 
tical in individuals with deciduous and per- 
manent antemolar dentitions. M. pusilla is 
notably smaller than M. jenkinsae in all ex- 
ternal, cranial, and dental measurements 
(Tables 1 and 2), and is separable based on 
pelage characters. 

A generic revision of all members of Mi- 
crogale is currently in preparation and in- 
cludes molecular characters (Olson and 
Goodman, in prep.). The results of this 
study will be presented elsewhere and will 
address aspects of the phylogenetic position 
and sister-taxa relationships of M. jenkin- 

Etymology. — This new species of Micro- 
gale is named after Paulina D. Jenkins of 
The Natural History Museum, London, for 
her important contributions to Tenrecidae 


Ecology. — Microgale jenkinsae is cur- 
rently known only from the Foret des Mi- 
kea, between Morombe and Manombo (Fig. 
1), in the southwestern portion of Mada- 
gascar, a zone of transitional dry deciduous 
and spiny bush habitat (Seddon et al. 2000; 
Goodman and Soarimalala in press). This 
region receives, on average, about 400-500 
mm of rainfall per year (Chaperon et al. 
1993), with probably more rainfall in the 
inland higher ground than along the coastal 
plain. Differences in forest types within the 
Foret des Mikea tend to follow this pattern, 
with more deciduous forest on the slightly 
higher ground away from the coast and 
spiny bush along the coastal plain. 



The climax vegetation of the dry decid- 
uous forest has been characterized as being 
dominated by the genera Dalbergia, Com- 
miphora, and Hildegardia (Humbert 1965). 
The formation has a canopy 10 to 15 m 
high, sometimes reaching 20 m, with a open 
medium stratum and diffuse undergrowth. 
All trees and most of the shrubs shed their 
leaves in the dry season. 

The vertebrate communities inhabiting 
the Foret des Mikea are similar to other arid 
portions of the island, although there are 
apparently several strict and regional en- 
demic vertebrates (e.g., the reptiles Furcifer 
belalandaensis and Paroedura vahiny, the 
birds JJratelornis chimaera and Monias 
benschi). The known small mammal com- 
munity consists of six Lipotyphla {Tenrec 
ecaudatus, Setifer setosus, Echinops telfai- 
ri, Geogale aurita, Microgale jenkinsae, 
and Suncus madagascariensis), one intro- 
duced murine rodent ( Rattus rattus), and 
two endemic Nesomyinae ( Macrotarsomys 
sp. and Eliurus myoxinus ) (Carleton and 
Schmidt 1990; Soarimalala and Goodman 
2004; Goodman and Soarimalala in press). 

Trapping . — During the survey of the 
Foret des Mikea, six sites were visited and 
systematically trapped using standard live 
and pit-fall traps (for more details on tech- 
niques see Goodman & Carleton 1996; 
Goodman et al. 1996). Pit-fall devices have 
been particularly useful for sampling Li- 
potyphla difficult to trap by other methods. 
For example, the only known specimens of 
another western Microgale, M. nasoloi, 
were captured with this technique (Jenkins 
and Goodman 1999). Three pit-fall lines, 
each composed of 11 12-liter buckets 
placed 10 m apart, were installed at each of 
the six survey sites. There was considerable 
variation between sites in trap success and 
species diversity. At site 1, 20 individuals 
( Tenrec , Setifer, Echinops, M. jenkinsae, 
and Geogale) were caught in 198 bucket 
nights between 14 and 19 February 2003. 
At site 2, 10 individuals ( Echinops and 
Geogale ) were obtained in 198 bucket 
nights between 21 and 27 February 2003. 

At site 3, 5 individuals ( Tenrec , Echinops, 
Rattus rattus, and Geogale) were taken in 
132 bucket nights between 2 and 5 March 
2003. At site 4, only 4 individuals ( Tenrec , 
Geogale, and Suncus madagascariensis) 
were trapped in 1 65 bucket nights between 
8 and 12 March 2003. At site 5, nothing 
was captured in 1 98 bucket nights between 
14 and 19 March 2003. At site 6, 7 indi- 
viduals ( Geogale ) were captured in 132 
bucket nights between 22 and 25 March 
2003. This level of faunal heterogeneity (as 
reflected by trapping-success) may be relat- 
ed to microhabitat differences between the 
sites, but further research is needed to test 
this hypothesis. 

Natural history. — Little definitive infor- 
mation can be gleaned on the natural his- 
tory of Microgale jenkinsae on the basis of 
two specimens, and the following extrapo- 
lations are tentative. Using foot structure 
and the context in which this species was 
trapped, it is terrestrial. Both specimens 
were taken in a portion of the Foret des 
Mikea dominated by more dry deciduous 
forest than by spiny bush habitat. Several 
of the other sites inventoried during the 
Foret des Mikea survey tended to be dom- 
inated by spiny bush habitat. The two M. 
jenkinsae were obtained in different pitfall 
lines installed in portions of the forest hav- 
ing a relatively dense understory — one line 
was dominated by Gramineae associated 
with the regeneration of an old forest ex- 
ploitation track and the other line in a mi- 
crohabitat with a dense growth of Xerophy- 
ta (Velloziaceae) mixed with other low- 
growing plants and some succulent Eu- 

An analysis of soil samples taken at each 
of the 18 pitfall lines installed during the 
Foret des Mikea survey indicate that at site 
1, where the two specimens of M. jenkinsae 
were collected, the average percent carbon 
in the soil was higher (1.9%, range 0.98- 
2.3%) than four of the other five sites (all 
less than 0.9% carbon). The outlier is site 
2 which had a slightly higher soil carbon 
content (2.4%, range 0.08-4.5%) than site 



1 . Given that shrew tenrecs are believed to 
be primarily insectivorous, one may expect 
that their distribution would be correlated 
with soils relatively rich in organic materi- 
al, which in turn would support a higher 
density and diversity of invertebrates. 

Both specimens are dentally sub-adults, 
and, thus, it is not unexpected that they did 
not show any signs of reproductive activity. 
However, shrew tenrecs with deciduous an- 
temolar dentitions can be reproductive (e.g., 
Jenkins et al. 1996). Our survey was con- 
ducted during the rainy season, a period of 
the year that normally coincides with breed- 
ing activity in small mammals in this area 
of the island (Ganzhorn et al. 1996; Ran- 
drianjafy 2003). 

Paleoecological implications. — Remains 
of Microgale pusilla have been reported in 
disintegrated owl pellets of unknown age, 
but by extrapolation almost certainly Hol- 
ocene, from the sites of Lelia and Anjo- 
himpaty in southwestern Madagascar 
(MacPhee 1986, 1987), a zone of xero- 
phytic spiny bush habitat on an exposed 
limestone substrate. M. pusilla is consid- 
ered to be an inhabitant of the more mesic 
portions of the island, including the eastern 
humid forest and central highlands. More 
recently bone remains of M. pusilla have 
been identified in owl pellets collected in 
the capital city of Antananarivo, at least 80 
km from the nearest intact natural forest 
block, and it is assumed that this species 
may live in surrounding marshlands and 
rice fields (Goodman et al. 1997a). At sev- 
eral sites in the central highlands it has been 
captured in marshlands within close vicinity 
to natural forest (Goodman et al. 2000a, 
Soarimalala et al. 2001). Thus, its occur- 
rence in owl pellets in southwestern Mad- 
agascar can be interpreted in at least two 
ways: this species is a generalist and is able 
to live in a variety of ecological conditions 
from humid forests to marshlands to xero- 
phytic bush — however, on the basis of re- 
cent inventories of the drier portions of the 
island there is no evidence of its occurrence 
in this latter habitat — or the undated owl 

pellets collected in the southwest are from 
a past geological period when this region of 
Madagascar was distinctly more mesic. 

The specimens of M. pusilla described by 
MacPhee (1986, 1987) from Lelia and An- 
johimpaty were deposited in the Service de 
Paleontologie collection at the Universite 
d’ Antananarivo. A detailed search of that 
collection, however, did not uncover these 
specimens. Nevertheless, a comparison of 
our material of M. jenkinsae to the illustra- 
tions and description of these specimens 
(MacPhee 1986) indicates considerable 
similarity in size, morphology, and dental 
structure. Most important in this regard is 
that dpm 2 (in M. jenkinsae and M. pusilla ) 
and pm 2 (in M. pusilla ) are simple in cor- 
onal structure and single-rooted, characters 
used to separate M. pusilla from all other 
small members of this genus before our rec- 
ognition of M. jenkinsae. Further, on the ba- 
sis of a scale provided with the line drawing 
of the Anjohimpaty mandible (MacPhee 
1986, Fig. 5), the approximate lower tooth- 
row length is 1 8.2 mm, which is within the 
range of M. jenkinsae, but notably larger 
than M. pusilla (Table 2). We strongly sus- 
pect that these specimens, reported as M. 
pusilla, may be referable to M. jenkinsae. 

Recent biological surveys of the Parc 
National de Tsimanampetsotsa (Fig. 1), for- 
merly under the statute of a Reserve Na- 
turelle Integrate, did not find any species of 
Microgale living in this protected area 
(Goodman et al. 2002), which is relatively 
close to Lelia and Anjohimpaty. Our small 
mammal surveys in the Foret des Mikea at 
six different sites, with a minimum of 132 
pit-fall nights per site, yielded a total of 
1023 pit-fall nights, yet only two individ- 
uals of M. jenkinsae were captured, both at 
the first site. This would indicate that this 
species is either rare or difficult to capture 
and presumably occupies specific micro- 
habitats. The important point here is that 
before significant paleoenvironmental infer- 
ences can be made associated with the pres- 
ence of certain taxa known only as subfos- 
sils, it is critical that detailed biological in- 



ventories be conducted in the general re- 
gion of the paleontological site to 
thoroughly document the extant fauna. 

Conservation. — Historically, most field 
efforts associated with the exploration and 
documentation of Madagascar’s unique fau- 
na have been in the humid forests on the 
eastern, central, and northern portions of 
the island. Further, there is a preponderance 
of reserves and parks protecting this biome 
as compared to the drier western and south- 
ern regions of the island (ANGAP 2001). 
On the basis of several recent biogeograph- 
ic analyses of small mammals and birds, 
species turnover along the nearly 1200 km 
long eastern humid forests of the island is 
relatively low (Goodman et al. 1997b, 
2000b). A number of endemic species in 
this biome have broad distributions, many 
extending the complete length of this hab- 
itat. More recent biological surveys of 
Madagascar’s western deciduous forests 
and southern spiny bush lands have re- 
vealed a previously unrecognized biota, in- 
cluding numerous terrestrial vertebrates. 
The growing realization is that levels of 
plant and animal species turnover along a 
latitudinal transect of western Madagascar 
is notably higher than the east, and this is 
probably related to a greater geological 
complexity and associated botanical com- 
munities in the west (Du Puy & Moat 

The recent surveys of the Foret des Mi- 
kea, which has no official protection, and 
forested regions to the north and south of 
this zone, are a case in point. Two unde- 
scribed species of mammals {Microgale 
jenkinsae and Macrotarsomys nov. sp.) 
have been discovered in the Foret des Mi- 
kea that are unknown from any other region 
in the west. Further to the north, in the vi- 
cinity of the Bemaraha Plateau, there are at 
least two species of rodents that appear to 
be endemic to the region (Carleton et al. 
2001). The recent discovery and description 
of Microgale nasoloi from unique forest 
formations in southwestern central Mada- 

gascar seems to indicate another regional 
endemic with a very limited distribution. 

The drier western and southern forests of 
Madagascar have been subjected to consid- 
erable anthropogenic degradation, perhaps 
greater than in the humid east (see Smith 
1997, Dufils 2003). In areas such as the 
Foret des Mikea, which was estimated in 
1999 to contain forest cover in excess of 
3700 km 2 , habitat loss rates have increased 
over the past few decades associated with 
pressures in the form of selective logging, 
cattle pasture, hunting, and clearing for ag- 
ricultural crops (Seddon et al. 2000). Given 
the levels of habitat heterogeneity and mi- 
croendemism in the west, action needs to 
be taken to protect the remaining large 
blocks of natural habitat in this region. On 
the basis of recent exploration of the Foret 
des Mikea this area should be given priority 
amongst the zones in need of rapid protec- 


We are grateful to the Direction des Eaux 
et Forets for issuing permits to conduct fau- 
nal surveys in the Foret des Mikea. For ac- 
cess to specimens in their care we are in- 
debted to Geraldine Veron, Museum Na- 
tional d’Histoire Naturelle, Paris; Paulina 
D. Jenkins, The Natural History Museum, 
London; and Prof. Daniel Rakotondravony, 
Universite d’ Antananarivo, Departement de 
Biologie Animale, Antananarivo. This field 
project was supported by WWF-Madagas- 
car, Fonds Frangaise pour l’Environnement 
Mondiale (1’Agence Frangaise de Devel- 
oppement), and the Volkswagen Founda- 
tion. We are grateful to Link Olson for 
comments on an earlier draft of this paper 
and his important aid in numerous ways. 
Two anonymous reviewers also helped to 
improve this paper. 

Literature Cited 

ANGAP (Association Nationale pour la Gestion des 
Aires Prot6g£es). 2001. Plan de gestion du re- 



seau national des aires protegees de Madagas- 
car. ANGAP, Antananarivo. 

Carleton, M. D., & D. F. Schmidt. 1990. Systematic 
studies of Madagascar’s endemic rodents (Mu- 
roidea: Nesomyinae): an annotated gazetteer of 
collecting localities of known forms.-American 
Museum Novitates 2987:1-36. 

, S. M. Goodman, & D. Rakotondravony. 2001. 

A new species of tufted-tail rat, genus Eliurus 
(Muridae: Nesomyinae), from western Mada- 
gascar, with notes on the distribution of E. 
myoxinus.— Proceedings of the Biological So- 
ciety of Washington 114:972-987. 

Chaperon, P., J. Danloux, & L. Ferry. 1993. Fleuves 
et rivieres de Madagascar. ORSTOM Editions, 

Dufils, J.-M. 2003. Remaining forest cover. Pp. 88-96 
in S. M. Goodman and J. P. Benstead, eds., The 
natural history of Madagascar. The University 
of Chicago Press, Chicago. 

Du Puy, D. J., & J. Moat. 2003. Using geological sub- 
strate to identify and map primary vegetation 
types in Madagascar and the implications for 
planning biodiversity conservation. Pp. 51-67 
in S. M. Goodman and J. P. Benstead, eds.. The 
natural history of Madagascar. The University 
of Chicago Press, Chicago. 

Ganzhorn, J. U., S. Sommer, J.-P. Abraham, M. Ade, 
B. M. Raharivololona, E. R. Rakotovao, C. 
Rakotondrasoa, & R. Randriamarosoa. 1996. 
Mammals of the Kirindy Forest with special 
emphasis on Hypogeomys antimena and the ef- 
fects of logging on the small mammal fauna. 
Pp. 215-232 in J. U. Ganzhorn and J.-P. Sorg, 
eds.. Ecology and economy of a tropical dry 
forest in Madagascar. — Primate Report 46- 1 . 

Goodman, S. M., & M. D. Carleton. 1996. The rodents 
of the Reserve Naturelle Integrale 
d’Andringitra, Madagascar. Pp. 257-283 in S. 
M. Goodman, ed., A floral and faunal inventory 
of the eastern slopes of the Reserve Naturelle 
Integrale d’Andringitra, Madagascar: with ref- 
erence to elevational variation. — Fieldiana: Zo- 
ology, new series 85. 

, & P. D. Jenkins. 1998. The insectivores of the 

Reserve Speciale d’Anjanaharibe-Sud, Mada- 
gascar. Pp. 139-161 in S. M. Goodman, ed., A 
floral and faunal inventory of the Reserve Spe- 
ciale d’Anjanaharibe-Sud, Madagascar: with 
reference to elevational variation. — Fieldiana: 
Zoology, new series 90. 

, , & O. Langrand. 1997a. Exceptional 

records of Microgale species (Insectivora: Ten- 
recidae) in vertebrate food remains. — Bonner 
Zoologische Beitrage 47:135-138. 

, , & D. Rakotondravony. 2000b. The 

biogeography of rodents (Rodentia: Muridae: 
Nesomyinae) and tenrecids (Lipotyphla: Tenre- 

cidae) in the eastern forests of Madagascar: an 
assessment of altitudinal zonation along a lati- 
tudinal gradient. Pp. 127-138 in W. R. Louren- 
50 and S. M. Goodman, eds., Diversite et en- 
demisme a Madagascar. Memoires de la Societe 
de Biogeographie, Paris. 

, M. Pidgeon, A. F. A. Hawkins, & T. S. Schu- 

lenberg. 1 997b. The birds of southeastern Mad- 
agascar. — Fieldiana: Zoology, new series 87:1 — 

, M. J. Raherilalao, D. Rakotomalala, D. Rak- 
otondravony, A. P. Raselimanana, H. V. Raza- 
karivony, & V. Soarimalala. 2002. Inventaire 
des vertebres du Parc National de Tsimanam- 
petsotsa (Toliara). — Akon’ny Ala 28:1-36. 

, D. Rakotondravony, M. J. Raherilalao, D. 

Rakotomalala, A. P. Raselimanana, V. Soari- 
malala, J.-M. Duplantier, J.-B. Duchemin, & J. 
Rafanomezantsoa. 2000a. Inventaire biologique 
de la foret de Tsinjoarivo, Ambatolampy. — 
Akon’ny Ala 27:18-27. 

, C. J. Raxworthy, & P. D. Jenkins. 1996. In- 

sectivore ecology in the Reserve Naturelle In- 
tegrale d’Andringitra, Madagascar. Pp. 218-230 
in S. M. Goodman, ed., A floral and faunal in- 
ventory of the eastern slopes of the Reserve Na- 
turelle Integrale d’Andringitra, Madagascar: 
with reference to elevational variation. — Field- 
iana: Zoology, new series 85. 

, & V. Soarimalala. in press. A new species of 

Macrotarsomys (Muridae: Nesomyinae) from 
the Foret des Mikea of southwestern Madagas- 
car — Proceedings of the Biological Society of 

Hershkovitz, P. 1977. Living New World monkeys 
(Platyrrhini) with an introduction to primates, 
vol. 1. The University of Chicago Press, Chi- 

Humbert, H. 1965. Description des types de vegeta- 
tion. Pp. 46-78 in H. Humbert and G. Cours- 
Darne, eds.. Notice de la carte de Madagascar. - 
Travaux de la Section Scientifique et Technique 
de FInstitut franqais de Pondichery. 

Jenkins, P. D. 1988. A new species of Microgale (In- 
sectivora: Tenrecidae) from northeastern Mad- 
agascar. — American Museum Novitates 2910: 

. 1 992. Description of a new species of Micro- 
gale (Insectivora: Tenrecidae) from eastern 
Madagascar. — Bulletin of the British Museum 
of Natural History, (Zoology) 58:53-59. 

. 1993. A new species of Microgale (Insecti- 
vora: Tenrecidae) from eastern Madagascar with 
an unusual dentition. — American Museum Nov- 
itates 3067:1-11. 

. 2003. Microgale, shrew tenrecs. Pp. 1273- 

1278 in S. M. Goodman and J. P. Benstead, eds., 



The natural history of Madagascar. The Uni- 
versity of Chicago Press, Chicago. 

, & S. M. Goodman. 1999. A new species of 

Microgale (Lipotyphla, Tenrecidae) from isolat- 
ed forest in southwestern Madagascar. — Bulle- 
tin of the Natural History Museum, London 
(Zoology) 65:155-164. 

, , & C. J. Raxworthy. 1996. The shrew 

tenrecs ( Microgale ) (Insectivora: Tenrecidae) of 
the Reserve Naturelle Integrate d’ Andringitra, 
Madagascar. Pp. 191-217 in S. M. Goodman, 
ed., A floral and faunal inventory of the eastern 
slopes of the Reserve Naturelle lntegrale 
d’Andringitra, Madagascar: with reference to 
elevational variation. — Fieldiana: Zoology, new 
series 85. 

— -*■ , C. J. Raxworthy, & R. A. Nussbaum. 1997. 
A new species of Microgale (Insectivora, Ten- 
recidae), with comments on the status of four 
other taxa of shrew tenrecs. — Bulletin of the 
Natural History Museum, London (Zoology) 

MacPhee, R.D.E. 1986. Environment, extinction, and 
Holocene vertebrate localities in southern Mad- 
agascar. — National Geographic Research 2: 

. 1987. The shrew tenrecs of Madagascar: sys- 
tematic revision and Holocene distribution of 
Microgale (Tenrecidae, Insectivora). — Ameri- 
can Museum Novitates 2889:1-45. 

Olson, L. E., & S. M. Goodman. 2003. Phylogeny and 
biogeography of tenrecs. Pp. 1235-1242 in S. 
M. Goodman and J. P. Benstead, eds.. The nat- 
ural history of Madagascar. The University of 
Chicago Press, Chicago. 

— , & A. D. Yoder, in press. Illumination 
of cryptic species boundaries in long-tailed 
shrew tenrecs (Mammalia: Tenrecidae; Micro- 
gale), with new insights into geographic varia- 
tion and distributional constraints. Biological 
Journal of the Linnean Society. 

Randrianjafy, R. V. 2003. Contribution a I’etude de 
biologie de conservation de la communaute mi- 
cromammalienne d’Ankarafantsika. These de 
Doctorat de 3eme cycle, Universite 
d’ Antananarivo, Antananarivo. 

Seddon, N„ J. Tobias, J. W. Yount, J. R. Ramanam- 
pamonjy, S. Butchart, & H. Randrianizahana. 
2000. Conservation issues and priorities in the 
Mikea Forest of south-west Madagascar. — Oryx 

Smith, A. P. 1997. Deforestation, fragmentation, and 
reserve design in western Madagascar. Pp. 415- 
441 in W. F. Laurance and R.O.W. Bierregaard 
Jr., eds.. Tropical forest remnants: ecology, 
management, and conservation of fragmented 
communities. The University of Chicago Press, 

Soarimalala, V. R. L., & S. M. Goodman. 2004. Les 
Rodentia, Lipotyphla et Carnivora de la foret 
des Mikea. Pp. 69-80 in A. P. Raselimanana 
and S. M. Goodman, eds., Inventaire floristique 
et faunistique de la foret de Mikea: Paysage 
ecologique et diversite biologique d’une pre- 
occupation majeure pour la conservation.— Re- 
cherches pour le Developpement, Serie Scienc- 
es biologiques. No. 21. 

Soarimalala, V., S. M. Goodman, H. Ramiarinjanahary, 
L. L. Fenohery, & W. Rakoronirina. 2001. Les 
micro-mammiferes non-volants du Parc Nation- 
al de Ranomafana et du couloir forestier qui la 
relie au Parc National d’Andringitra. Pp. 197- 
229 in S. M. Goodman and V. R. Razafindrat- 
sita, eds., Inventaire biologique du Parc Nation- 
al de Ranomafana et du couloir forestier qui la 
relie au Parc National d’Andringitra. Recherch- 
es pour le Developpement, Serie Sciences Biol- 
ogiques, Centre d’lnformation et de Documen- 
tation Scientifique et Technique, Antananarivo, 
no. 17. 

Appendix 1 

List of specimens of Microgale spp. examined during 
the course of this study. 

Microgale parvula . — Province d’Antsiranana, Parc 
National de Marojejy [formerly Reserve Naturelle In- 
tegrate de Marojejy], along tributary of Manantenina 
River, 8 km NW Manantenina, 14°26.2'S, 49°46.5'E, 
450 m (FMNH 159681); Parc National de Marojejy 
[formerly Reserve Naturelle Integrate de Marojejy], 
along tributary of Manantenina River, 10 km NW 
Manantenina, 14°26.0'S, 49°45.7'E, 775 m (FMNH 
159682, 159683, 159684); Parc National de Marojejy 
[formerly Reserve Naturelle Integrate de Marojejy], 1 1 
km NW Manantenina, Antranohofa, 14°26.2'S, 
49°44.5'E, 1225 m (FMNH 159685); Parc National de 
Marojejy [formerly Reserve Naturelle Integrate de Ma- 
rojejy], 10.5 km NW Manantenina, along tributary at 
head of Andranomifototra River, 14°26.4'S, 49°44.5'E, 
1625 m (FMNH 159686). Province de Fianarantsoa, 
approx. 45 km S. Ambalavao, east bank Iantara River, 
along Ambalamanenjana-Ambatoboay trail, edge of 
Parc National d’Andringitra [formerly Reserve Natu- 
relle Integrate], 22°13'20"S, 47°0I'29"E, 720 m 
(FMNH 151621); Parc National d’Andringitra [for- 
merly Reserve Naturelle Integrate d’Andringitra], 
approx. 43 km S. Ambalavao, junction of Sahanivo- 
raky and Sahavatoy Rivers, 22°13'40"S, 47°00'13''E, 
810 m (FMNH 151622); Parc National d’ Andringitra 
[formerly Reserve Naturelle Integrate d’Andringitra], 
approx. 38 km S. Ambalavao, on ridge east of Volot- 
sangana River, 22°11'39"S, 46°58'16"E, 1625 m 
(FMNH 151623, 151723, 151794, 151801). 

Microgale nasoloi . — Province de Toliara, Foret de 
Vohibasia, 59 km northeast Sakaraha, 780 m. 



22°27.5'S, 44°50.5'E (FMNH 156187); Foret 

d’Analavelona, Antanimena, 12 .5 km NW Andranoh- 
eza, 22°40.7'S, 44°11.5'E, 1050 m (FMNH 161576). 

Microgale fotsifotsy . — Province de Fianarantsoa, 
Parc National d’Andringitra, 8.5 km SE Antanifotsy, 
Campement Andohan’ Ambola, 22°10.273'S, 
46°56.758'E, 1960 m (FMNH 165694, 165778, 
165779) ; 2 km W. Andrambovato, along Tatamaly 
River, 21°30.7'S, 47°24.6'E, 1075 m (FMNH 170749); 
Foret de Vinantelo, at foot of Mt. Ambodivohitra, 15.5 
km SE Vohitrafeno, 2I°46.6'S, 47°20.8'E, 1100 m 
(FMNH 170750); approx. 40 km S. Ambalavao, along 
Volotsangana River, 22°13'22"S, 46°58'18"E, 1210 m 
(FMNH 151646, 151647); Province de Mahajanga, 
western side of Anjanaharibe-Sud, 13.5 km SW Befin- 
gotra, 14°47.0'S, 49°26.5'E, 1200 m (FMNH 167428). 
Province de Toliara, Parc National d’Andohahela [for- 
merly Reserve Naturelle Integrale d’Andohahela], par- 

cel I, 8 km NW Eminiminy, 24°37.55'S, 46°45.92'E, 
440 m (FMNH 156569); Parc National d’Andohahela 
[formerly Reserve Naturelle Integrale d’Andohahela], 
parcel I, 13.5 km NW Eminiminy, 24°35.04'S, 
46°44.08'E, 1200 m (FMNH 156424); Parc National 
d’Andohahela [formerly Reserve Naturelle Integrale 
d’Andohahela], parcel I, 15.0 km NW Eminiminy, 
24°35.15'S, 46°43.85'E, 1500 m (FMNH 156570). 

Microgale pusilla . — Province d’ Antananarivo, 13 
km NE Antananarivo, in Tyto alba pellets (FMNH 
151606, 151607); 10 km SE Tsinjoarivo, Foret de Ma- 
hatsinjo, Andasivodihazo, 19°40.7'S, 47°46.2'E, 1550 
m (FMNH 166123, 166124, 166125) ; Reserve Spe- 
ciale d’Ambohitantely, 24 km NE Ankazobe, 
18°10.1'S, 47°16.6'E, 1450 m (FMNH 165489). Prov- 
ince de Fianarantsoa, Manambolo Forest, Ambavafa- 
tra, along Andohabatotany River, 17.5 km SE Sendri- 
soa, 22°8'58"S, 47°1'25", 1300 m (FMNH 167612, 
167619, 167621).